Researcher develops personalized search engines

May 15, 2012

(Phys.org) -- With little more than basic information about Web users’ behavior – that is, the hyperlinks they click on daily and the content at those sites – Susan Gauch can build a better search engine. In information systems research, this work is known as “implicit” user profiling, meaning there are basic assumptions about user interest and intent based on the sites they frequent and the content they view.

Gauch, a professor of computer science and computer engineering at the University of Arkansas, has expertise in developing robust and personalized search engines, which she will contribute to the work of Hypothes.is, a project started by Dan Whaley, the coder and entrepreneur who built the first Web-based travel reservation system. Hypothes.is is trying to build a system of annotation for the Web. Based on a model of community peer-review, the system will be an open-source platform that will enable annotators to comment on individual sentences.

“Since the very beginning of the Web, there has been an issue of trust,” Gauch said, “because there has always been this ubiquitous ability for anyone to create and distribute information. What Hypothes.is is trying to do is build confidence and trust about information obtained on the Web. Yes, it is a form of peer review, but it won’t be hierarchical or purely academic. Many details haven’t been worked out yet, but the peer-review component will be determined by the annotator’s reputation, which will be based on many demographic factors and will be constantly under review by other annotators.”

The Hypothes.is system will function as an overlay – similar to the “Track Changes” tool in Microsoft Word – on top of stable content, including news, blogs, scientific articles, books, terms of service, ballot initiatives, legislation and regulations, software code and more. As an overlay, the system will not require participation of the underlying site, and the content will reside on separate servers, Gauch said. She added that by enabling or disabling plug-ins, users will have the power to turn the service on and off.

The work of Hypothes.is will build on the unique expertise of people like Gauch. Her work in user profiling relies on queries and log-in applications, proxy servers and especially plug-ins and cookies. Plug-ins can be thought of as bonus features attached to primary or master software. They are extra components that add specific abilities. A computer cookie is a small text file with a unique identification tag placed on the user’s computer by a website. Among other purposes, these features allow researchers to track user behavior. Gauch’s task then is to take this information and build a personalized search.

“If we know something about you, then we can build a profile and use different functions to tailor searches specific to your interests,” Gauch said. “Technology allows us to see which sites users frequently visit, so we have a good idea what they’re looking for when they enter vague or ambiguous search terms.”

How does this affect the user? Gauch mentions various search terms that cause great confusion without at least some basic knowledge about the user’s interests. For example, consider the word “rock” as a search term. With no information about the user’s browsing history, a simple Google search finds sites related to gems, geology, music (bands, merchandise and the hall of fame and museum in Cleveland), wrestling, the actor (Dwayne Johnson), political action groups (Rock the Vote), county government (Rock County, Minnesota, and Rock County, Wisconsin) and something called “Rock Your Phone.” The information about a user’s behavior allows Gauch to build profiles that so that musicians will be directed to sites that sell guitars, and geologists will find the U.S. Geological Survey site.

Explore further: Computerized emotion detector

add to favorites email to friend print save as pdf

Related Stories

iSchool prof predicts the future of search user interfaces

Nov 07, 2011

School of Information professor Marti Hearst predicts the future of online search interfaces in an article in this month’s edition of the Communications of the ACM. “The future of user interfaces will involv ...

Improving web search

Jan 16, 2012

Research from Victoria University could help search engines understand people’s queries much better.

The influence of the internet on suicidal behavior

Oct 10, 2011

A recent study by the University of Otago, Wellington into internet pro-suicide and support sites indicates that significant improvements need to be made in this area to help prevent suicidal behavior.

Safe 'sandbox' for the internet of the future

May 13, 2011

To better protect new Internet applications against hacker attacks and other types of manipulation, Siemens is taking part in the EU’s WebSand research project. In cooperation with partner organizations ...

Recommended for you

Computerized emotion detector

Sep 16, 2014

Face recognition software measures various parameters in a mug shot, such as the distance between the person's eyes, the height from lip to top of their nose and various other metrics and then compares it with photos of people ...

Cutting the cloud computing carbon cost

Sep 12, 2014

Cloud computing involves displacing data storage and processing from the user's computer on to remote servers. It can provide users with more storage space and computing power that they can then access from anywhere in the ...

Teaching computers the nuances of human conversation

Sep 12, 2014

Computer scientists have successfully developed programs to recognize spoken language, as in automated phone systems that respond to voice prompts and voice-activated assistants like Apple's Siri.

Mapping the connections between diverse sets of data

Sep 12, 2014

What is a map? Most often, it's a visual tool used to demonstrate the relationship between multiple places in geographic space. They're useful because you can look at one and very quickly pick up on the general ...

User comments : 0